This repository provides a Reconciliation Service for the the SkoHub core infrastructure.

Dependencies:

• elasticsearch 7.0
• node-version >= v12.16.1

Basic setup:

$ git clone https://github.com/rg-mpg-de/skohub-reconcile.git
$ cd skohub-reconcile
# Copy `sample.env` to `.env` and adjust values therein
$ npm ci
$ npm test
$ npm start

This will start the Reconciliation service on the port specified with APP_PORT in .env. It accepts queries according to the Reconciliation Service specification at endpoints corresponding to all hosted vocabularies, e.g. /project/vocab, /class/esc, or /rg-mpg-de/polmat etc.

Currently only reconciliation queries are supported. Preview, suggestion and data extensions support is on the roadmap.

The elasticsearch server must be populated when a vocabulary is published on skohub. This present service creates an appropriate index and takes PUT requests from skohub-vocabs, adding resources to the elasticsearch index.

You need to run a properly configured elasticsearch instance by setting cluster.name: skohub. See the provided elasticsearch.yml. Also, in some contexts, it's mandatory to initialize elasticsearch with a proper index-mapping.

You may want to use the start script in scripts/start.sh. This script ensures the proper installation of skohub-reconcile and the configuration of elasticsearch. There also reside further scripts to manage the starting/stopping of the skohub-reconcile via init and to monitor the processes with monit.

The project to add a Reconciliation Service to SkoHub has been initiated by Andreas Wagner and carried out in cooperation with the SkoHub.io team in 2021/2022.